Systems and methods for mapping in-store transactions to customer profiles

Abstract

A system including one or more processors and one or more non-transitory computer-readable media storing computing instructions configured to run on the one or more processors and perform: receiving a query from a front-end device for one or more users mapped to a same payment option; generating, using a machine learning model, a first dataset comprising one or more classifications of one or more online users mapped to the same payment option as either (i) a single user with multiple user profiles or (ii) multiple users of a same household; generating, using a factor graph, a second dataset comprising first information of the one or more online users mapped to second information of one or more instore users; mapping at least one of the one or more online users to at least one of the one or more instore users based on the second dataset. Other embodiments are disclosed.

What is claimed is: 
     
       1. A system comprising:
 one or more processors; and 
 one or more non-transitory computer-readable media storing computing instructions configured to run on the one or more processors and perform:
 receiving, at a back-end device, a query from a front-end device for one or more users mapped to a same payment option; 
 generating, using a machine learning model, a first dataset comprising one or more classifications of one or more online users mapped to the same payment option as either (i) a single user with multiple user profiles or (ii) multiple users of a same household; 
 generating, using a factor graph, a second dataset comprising first information of the one or more online users mapped to second information of one or more instore users, wherein the first information comprises a plurality of types of attributes of one or more user profiles of the one or more online users, and wherein the second information of the one or more instore users comprises a plurality of instore transactions; 
 recording contact information and biometric information of an instore user of the one or more instore users by a point-of-sale terminal at a store location, wherein the contact information and the biometric information is added as an observed variable of observed variables on the factor graph, wherein the biometric information comprises an image of the instore user, and wherein the contact information is linked to the second dataset; 
 mapping at least one of the one or more online users to at least one of the one or more instore users based on the second dataset; 
 generating a third dataset by combining the first dataset and the second dataset, wherein the third dataset comprises the at least one of the one or more online users linked to the at least one of the one or more instore users; 
 creating a unified view of each user of the one or more users based on transactions combined in the third dataset; and 
 sending instructions to display the third dataset to the front-end device. 
 
 
     
     
       2. The system of  claim 1 , wherein the computing instructions are further configured to perform:
 training a machine learning model based on historical online transactions of the one or more online users, wherein input data for the machine learning model comprises the one or more user profiles of the one or more online users and one or more payment options used during a period of time, and output data for the machine learning model comprises the one or more classifications. 
 
     
     
       3. The system of  claim 1 , wherein the machine learning model comprises a logistic regression model. 
     
     
       4. The system of  claim 1 , wherein the computing instructions are further configured to perform:
 determining one or more observed variable nodes and one or more unobserved variable nodes of the factor graph that have a shared usage of at least one same payment option, wherein the one or more user profiles are mapped to the same payment option; 
 creating dependences between the observed variables and unobserved variables of the factor graph based on an overlap of the shared usage of the at least one same payment option, wherein the factor graph models the dependencies between the observed variables and the unobserved variables in a probabilistic graphical model; and 
 encoding logic into a function node of function nodes on the factor graph, wherein the logic assigns connections between the observed variable nodes and the unobserved variable nodes of the factor graph. 
 
     
     
       5. The system of  claim 1 , wherein determining the second dataset further comprises:
 identifying, using the machine learning model, a respective gender of each image of the instore user captured for each respective user identification of a subset of user identifications. 
 
     
     
       6. The system of  claim 1 , wherein determining the second dataset further comprises:
 identifying one or more similarities between (a) a second type of attribute of the plurality of types of attributes in the first information and (b) the second information of the one or more instore users, 
 wherein mapping the at least one of the one or more online users to at least one of the one or more instore users further comprises:
 mapping at least one user profile of the one or more user profiles to at least one instore user of the one or more instore users, wherein the at least one user profile comprises the contact information of the instore user. 
 
 
     
     
       7. The system of  claim 1 , wherein determining the second dataset further comprises:
 calculating weights associated with the one or more user profiles, wherein the weights are calculated based on one or more similarities between the plurality of types of attributes associated with the one or more user profiles. 
 
     
     
       8. The system of  claim 1 , wherein the computing instructions are further configured to perform:
 receiving third information indicating that unobserved variable nodes on the factor graph are related; 
 extending the factor graph to map the unobserved variable nodes to at least one same payment option based at least in part on the third information; and 
 determining, by logic, that the one or more user profiles are associated with the one or more instore users of a same household. 
 
     
     
       9. The system of  claim 1 , wherein the computing instructions are further configured to perform:
 receiving multiple user identifications of the multiple users and respective payment options of multiple payment options in association with multiple transactions, wherein the one or more user profiles comprise the multiple user identifications associated with the multiple users, and wherein the multiple payment options comprise one or more same payment options; 
 distilling a subset of payment options from among the multiple payment options; 
 distilling a subset of user identifications from among the multiple user identifications, such that each respective payment option of the subset of payment options is associated with more than one user identification of the subset of user identifications; 
 analyzing the subset of payment options to create an analysis of patterns based on the subset of user identifications associated with a first single payment option of the subset of payment options; 
 identifying, using the analysis, a first classification of the one or more classifications comprising the single user of the subset of user identifications that is mapped to more than one user identification of the subset of user identifications; and 
 identifying, using the analysis, a second classification of the one or more classifications comprising two or more first users of the subset of user identifications that are part of the same household. 
 
     
     
       10. The system of  claim 9 , wherein the computing instructions are further configured to perform:
 filtering each respective user identification of the subset of user identifications, comprising:
 determining whether a record is classified in the first classification by determining a respective first similarity score for each of one or more respective multiple user identifications of the subset of user identifications mapped to a same payment option; 
 determining whether the record is classified in the second classification by determining similarities in user information; and 
 filtering the similarities using a second similarity score, wherein the determining whether the record is classified in the second classification further comprises:
 searching the user information comprising at least first names, last names, and physical addresses; 
 applying an edit distance on the first names and the last names between all pairs of the user information in the record that shared a second single payment option; 
 filtering the user information by the second single payment option attributable to the multiple user identifications and by transaction history; and 
 clustering the user information into the second classification when the edit distance for the user information falls below a preset threshold, 
 
 
 wherein the respective first similarity score for each of the one or more respective multiple user identifications of the subset of user identifications is indicative of similarities between a string of one or more pieces of the user information associated with a respective one of the multiple user identifications and a respective string of the one or more pieces of the user information associated with each respective one or more other ones of the multiple user identifications. 
 
     
     
       11. A method being implemented via execution of computing instructions configured to run at one or more processors and stored at one or more non-transitory computer-readable media, the method comprising:
 receiving, at a back-end device, a query from a front-end device for one or more users mapped to a same payment option; 
 generating, using a machine learning model, a first dataset comprising one or more classifications of one or more online users mapped to the same payment option as either (i) a single user with multiple user profiles or (ii) multiple users of a same household; 
 generating, using a factor graph, a second dataset comprising first information of the one or more online users mapped to second information of one or more instore users, wherein the first information comprises a plurality of types of attributes of one or more user profiles of the one or more online users, and wherein the second information of the one or more instore users comprises a plurality of instore transactions; 
 recording contact information and biometric information of an instore user of the one or more instore users by a point-of-sale terminal at a store location, wherein the contact information and the biometric information is added as an observed variable of observed variables on the factor graph, wherein the biometric information comprises an image of the instore user, and wherein the contact information is linked to the second dataset; 
 mapping at least one of the one or more online users to at least one of the one or more instore users based on the second dataset; 
 generating a third dataset by combining the first dataset and the second dataset, wherein the third dataset comprises the at least one of the one or more online users linked to the at least one of the one or more instore users; 
 creating a unified view of each user of the one or more users based on transactions combined in the third dataset; and 
 sending instructions to display the third dataset to the front-end device. 
 
     
     
       12. The method of  claim 11 , further comprising:
 training a machine learning model based on historical online transactions of the one or more online users, wherein input data for the machine learning model comprises the one or more user profiles of the one or more online users and one or more payment options used during a period of time, and output data for the machine learning model comprises the one or more classifications. 
 
     
     
       13. The method of  claim 11 , wherein the machine learning model comprises a logistic regression model. 
     
     
       14. The method of  claim 11 , further comprising:
 determining one or more observed variable nodes and one or more unobserved variable nodes of the factor graph that have a shared usage of at least one same payment option, wherein the one or more user profiles are mapped to the same payment option; 
 creating dependences between the observed variables and unobserved variables of the factor graph based on an overlap of the shared usage of the at least one same payment option, wherein the factor graph models the dependencies between the observed variables and the unobserved variables in a probabilistic graphical model; and 
 encoding logic into a function node of function nodes on the factor graph, wherein the logic assigns connections between the observed variable nodes and the unobserved variable nodes of the factor graph. 
 
     
     
       15. The method of  claim 11 , wherein determining the second dataset further comprises:
 identifying, using the machine learning model, a respective gender of each image of the instore user captured for each respective user identification of a subset of user identifications. 
 
     
     
       16. The method of  claim 11 , wherein determining the second dataset further comprises:
 identifying one or more similarities between (a) a second type of attribute of the plurality of types of attributes in the first information and (b) the second information of the one or more instore users, 
 wherein mapping the at least one of the one or more online users to at least one of the one or more instore users further comprises:
 mapping at least one user profile of the one or more user profiles to at least one instore user of the one or more instore users, wherein the at least one user profile comprises the contact information of the instore user. 
 
 
     
     
       17. The method of  claim 11 , wherein determining the second dataset further comprises:
 calculating weights associated with the one or more user profiles, wherein the weights are calculated based on one or more similarities between the plurality of types of attributes associated with the one or more user profiles. 
 
     
     
       18. The method of  claim 11 , further comprising:
 receiving third information indicating that unobserved variable nodes on the factor graph are related; 
 extending the factor graph to map the unobserved variable nodes to at least one same payment option based at least in part on the third information; and 
 determining, by logic, that the one or more user profiles are associated with the one or more instore users of a same household. 
 
     
     
       19. The method of  claim 11 , further comprising:
 receiving multiple user identifications of the multiple users and respective payment options of multiple payment options in association with multiple transactions, wherein the one or more user profiles comprise the multiple user identifications associated with the multiple users, and wherein the multiple payment options comprise one or more same payment options; 
 distilling a subset of payment options from among the multiple payment options; 
 distilling a subset of user identifications from among the multiple user identifications, such that each respective payment option of the subset of payment options is associated with more than one user identification of the subset of user identifications; 
 analyzing the subset of payment options to create an analysis of patterns based on the subset of user identifications associated with a first single payment option of the subset of payment options; 
 identifying, using the analysis, a first classification of the one or more classifications comprising the single user of the subset of user identifications that is mapped to more than one user identification of the subset of user identifications; and 
 identifying, using the analysis, a second classification of the one or more classifications comprising two or more first users of the subset of user identifications that are part of the same household. 
 
     
     
       20. The method of  claim 19 , further comprising:
 filtering each respective user identification of the subset of user identifications, comprising:
 determining whether a record is classified in the first classification by determining a respective first similarity score for each of one or more respective multiple user identifications of the subset of user identifications mapped to a same payment option; 
 determining whether the record is classified in the second classification by determining similarities in user information; 
 filtering the similarities using a second similarity score, wherein the determining whether the record is classified in the second classification further comprises:
 searching the user information comprising at least first names, last names, and physical addresses; 
 applying an edit distance on the first names and the last names between all pairs of the user information in the record that shared a second single payment option; 
 filtering the user information by the second single payment option attributable to the multiple user identifications and by transaction history; and 
 clustering the user information into the second classification when the edit distance for the user information falls below a preset threshold, 
 wherein the respective first similarity score for each of the one or more respective multiple user identifications of the subset of user identifications is indicative of similarities between a string of one or more pieces of the user information associated with a respective one of the multiple user identifications and a respective string of the one or more pieces of the user information associated with each respective one or more other ones of the multiple user identifications.